In this research, the emotions and the patterns of EEG signals of human brain are studied. The aim of this research is to study the analysis of the changes in the brain signals in the domain of different emotions. The observations can be analysed for its utility in the diagnosis of psychosomatic disorders like anxiety and depression in economical way with higher precision.
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Analysis of emotion disorders based on EEG signals ofHuman Brain
1. International Journal of Computer Science, Engineering and Applications (IJCSEA) Vol.2, No.4, August 2012
DOI : 10.5121/ijcsea.2012.2403 19
Analysis of emotion disorders based on EEG
signals of Human Brain
Ashish Panat 1
and Anita Patil 2
1
Priyadarshini Indira College of Engineering
asishpanat@gmail.com
2
Department of Electronics & Telecommunication, Cummins College of Engineering for
Women, Pune.
anita.patil&@cumminscollege.in
ABSTRACT
In this research, the emotions and the patterns of EEG signals of human brain are studied. The aim of this
research is to study the analysis of the changes in the brain signals in the domain of different emotions. The
observations can be analysed for its utility in the diagnosis of psychosomatic disorders like anxiety and
depression in economical way with higher precision.
KEYWORDS
EEG, EDF format, Feature extraction, Image classifiers, Emotions, Psychosomatic disorders, Normal and excited
brain, Anxiety and Depression
1. INTRODUCTION
A Human Brain is the organ that gives the person the capacity for art, language, rational thoughts
and moral judgments. It is also responsible for each individual's personality, movements,
memories, and his perception about the world. It is one of the body's biggest organs, consisting of
some 100 billion nerve cells that not only put together and highly coordinated physical actions but
regulate our unconscious body processes, such as digestion and breathing.
Emotions play a significant and powerful role in everyday life of human beings. Impulsive
emotions express an indication of psychosomatic disorders. These disorders are reflected as the
changes in the electrical activities and chemical activities in the brain. The changes can be
observed by capturing the brain signals and images.
Psychiatrists nowadays, have to deal with the patients with either of two prominent psychological
disorders, viz., Anxiety and Depression. Moreover, the patients are not ready to accept that the
symptoms they are suffering from are indicative of some psychological disorder. It becomes a
difficult job for the Psychiatrist, relatives of the patients and people around him to convince that
he needs to be treated.
The proposed research is expected to quantify the psychological health of the patient from his
EEG, as far as the two problems mentioned above, i.e. Anxiety and Depression are considered.
2. International Journal of Computer Science, Engineering and Applications (IJCSEA) Vol.2, No.4, August 2012
20
Anxiety disorder: The term anxiety disorder covers several different forms of abnormal and
pathological fear and anxiety. It covers four aspects of experiences an individual may have:
mental apprehension, physical tension, physical symptoms and dissociative anxiety [1]. Anxiety
disorder is divided into three types viz., anxiety disorder, phobic disorder, and panic disorder;
each has its own characteristics and symptoms. They also require different treatment. The
emotions present in anxiety disorders range from simple nervousness to bouts of terror. The
amygdala is central to the processing of fear and anxiety, and its function may be disrupted in
anxiety disorders.
Anxiety disorder is a pattern of constant worry and anxiety over many different activities and
events. Other symptoms include difficulty in concentrating, fatigue, irritability, restlessness, and
often becoming startled very easily [2]. A phobic disorder is a persistent fear of an object or
situation. Panic disorder is an anxiety disorder characterized by recurring severe panic attacks. It
may also include significant behavioral changes lasting long and ongoing worry about the
implications or concern[3].
Depression: Depression is a state of low mood and aversion to activity that can affect a person's
thoughts, behaviour, feelings and physical well-being [4]. Depressed people may feel sad,
anxious, empty, worthless, guilty, irritable, or restless. They may lose interest in activities that
once were pleasurable, experience loss of appetite or overeating, or problems concentrating,
remembering details or making decisions. They may even contemplate or attempt suicide.
Insomnia, excessive sleeping, fatigue, loss of energy, or aches, pains or digestive problems that
are resistant to treatment may be present. The signals of the brain in these situations can also be
utilized to study the emotions which can lead to great help in diagnosis of psychosomatic
disorders.
The research is conducted previously to analyse the emotions by looking at the physiological
aspects like users’ heart rate, skin conductance and pupil dilation.
2. LITERATURE SURVEY
In [5] & [6], Researchers Z. Khalili et al. have worked on Emotion detection using EEG and
peripherals signals as Galvanic skin resistance, Respiration, Blood pressure, and Temperature.
From these inputs, common set of features such as Mean, standard deviation and minimum and
maximum of the set of data are extracted. Their research is further extended to study the
improvement in the results of EEG by using correlation dimension.
In [7], the same researchers have explored on different modalities for emotion detection, such as,
Visual (facial expression), Auditory (pitch, loudness, etc.), tactile (heart rate, skin conductivity
etc.) and Brain signals(EEG).
In [8], Researchers have studied Brain activation during judgments of Positive emotions: Pride
and Joy. They have used fMRI images for this purpose. However, recording fMRI of a brain is
comparatively costly affair as compared to recording of EEG.
Researchers Arman Savran et al. [9] have also developed a technique for multimodal emotion
detection. They have used the modalities like fNIRS, face video and EEG signal.
3. BLOCK DIAGRAM
Figure(1) shows the block diagram of the system to acquire the signals from the brain, pre-
processing of the captured signals ( e.g. EEG in this study), extract the features after processing
the signal, classify the processed signal and analyse it for detection of emotion disorder.
3. International Journal of Computer Science, Engineering and Applications (IJCSEA) Vol.2, No.4, August 2012
21
Fig.(1) Block Diagram of the system for analysis of Brain signals.
Initially, the EEG of the Brain is captured by the standard method recognized worldwide as
International 10-20 system. In the Digital EEG system, these signals are first amplified and then
digitized. The rate of digitization may vary from 100 Hz to 20 kHz, depending on the capacity of
the system. Most commonly, the EEG signal captured from the EEG machine is available in the
EDF format (European Data Format). It must be first converted into .Wav format which is
suitable for processing. The signals are then filtered. The pass band of the filter depends on the
frequency of the interest for that particular signal, e.g. a low pass filter for Delta and Theta waves,
a band pass filter for alpha waves, and a high pass filter for gamma waves etc.
After filtering the signal, the features can be extracted, which can be compared and used for
further analysis.
4. CAPTURING THE SIGNAL
The fig.(2) shows the placement of the electrodes for capturing the EEG.
Fig (2) Placement of electrodes for EEG [12]
As per the International Standard, known as international 10-20 system, 19 electrodes are
connected to different locations on the scalp, which are salient points from the clinical point of
view, and one reference electrode is connected to ground. Whenever, the detailed study of EEG is
intended in case of some patients, or for the research purpose, the number of electrodes may
increase up to 256 also. In case of infants, i.e. neonatal EEG, the number can be decreased.
5. PREPROCESSING
The EEG signal from the EEG machine is available in the EDF format (European Data Format).
It is first converted into .Wav format which is suitable for processing. The software EDF2WAV is
used for this purpose. These WAV files can be directly referred into MATLAB programs, as
input data. The different frequency bands. The Fig. (3) shows the flowchart of different steps of
Pre-processing followed by feature extraction.
4. International Journal of Computer Science, Engineering and Applications (IJCSEA) Vol.2, No.4, August 2012
22
Fig.(3) Pre-processing and feature extraction
6. FEATURE EXTRACTION
After pre- processing, the following features are extracted from this signal: Mean, Standard
deviation, skewness, kurtosis, mean of absolute values of first difference of raw signals, mean of
absolute values of first difference of normalized signal [5]. The images of these signals can be
stored for further analysis. For feature extraction, basic mathematical formulae for mean, variance
etc. can be applied. Also, one can use the wavelet transform for this purpose.
Comparison of these features can be made to find the emotion disorder. As a first step towards
detection of emotional disorder, one obvious symptom, seizure, is identified by comparing EEGs
of a patient in normal condition & with seizure.
Seizure is one of the symptoms that may occur in a patient suffering from anxiety. The abnormal
activity of the brain in epileptic patient is known as seizure condition. This activity appears on the
screen of the EEG machine as waveforms of varying frequency and amplitude measured in
voltage. In this study, two features, viz., Mean and Variance are compared to detect the seizure
occurred in the patient.
7. RESULTS
The following snap shots from MATLAB result windows show the conversion of EDF file
(Fig.4) to .WAV file (Fig.5), band pass filter applied to theta wave Fig.(6).
Fig.(4) Signal extracted from .EDF file.
Fig.(5) Result of MATLAB code for EEG in .WAV format
Fig(6) Result of Band pass filter for Theta wave (using FDA tool)
5. International Journal of Computer Science, Engineering and Applications (IJCSEA) Vol.2, No.4, August 2012
23
: Normal : seizure
Fig.(7) Plot of Mean for Normal and seizure condition
: Normal : seizure
Fig.(8) Plot of Variance from EEG signal recorded for Normal and seizure condition
Result Table
Feature Accuracy
Mean 75 %
Variance 93%
Table 1: Comparison of results of two features, Mean and Variance
8. CONCLUSION
The study has proved the effective utility of economical and simple method of study of brain
using EEG for diagnosis the different emotion disorders viz., anxiety and depression. One of the
major symptoms, seizure is analysed in this study. The EEG signal in EDF format is converted
into .WAV format using EDF to WAV converter. The signal is then passed through the filters of
different frequencies to separate alpha, beta, delta and theta waves. The features are extracted
from these signals. After analyzing Mean, Power Spectral Entropy and Variance of both normal
and seizer EEG signals, it was found that using Variance gives more accurate information about
the seizure in EEG signal among the given feature sets.
This technique has revealed the possibility of precise diagnosis of psychosomatic disorders in
more simple and economical way.